基于冲突因子和向量余弦的迭代D-S加权算法

摘要

为了扩大D-S方法的适用范围,用传统的冲突因子与Pignistic概率向量余弦共同度量证据间的矛盾程度,将矛盾系数转化为证据的权系数,首轮融合时只用权系数修正证据集,其他融合用替换修正和权系数修正二次修正证据集,采用迭代思想修正融合结果.替换修正充分利用每一次的融合结果,迭代过程使得最终的融合结果更加精确.利用公共算例和随机算例验证后表明所提方法的融合效率高,融合结果更准确,算法收敛速度快,能有效解决问题,扩大了D-S的使用范围.%In order to expand the application scope of D-S evidence theory,the conflict degree is measured with traditional conflict factor and Pignistic probability vector cosine to convert the conflict coefficient into the weight coefficient of evidence. The weight coefficient is used to modify the evidence set for the first fusion. The replace correction and weight coefficient correc-tion are used to secondly modify the evidence set for other fusions. The iteration thought is adopted to modify the fusion result. Every fusion result is fully utilized by replace correction to make the final fusion result more accurate in iteration process. The verification results of classical example and random example show this algorithm has high fusion efficiency,accurate fusion re-sult and fast convergence speed. It solves the problem effectively and extends the application range of D-S method.